This project describes a biochemical engineering approach to design a process and platform for the de novo synthesis of large DNA molecules from precursor synthetic 5'OH oligonucleotides, 30-40 nucleotides long. The instrument is engineered with two applications in mind: (1) assembly of genes from synthetic oligonucleotides and (2) site-directed mutagenesis. The unique features of the technology are the speed of assembly and the low error frequencies of products. The overall objective (end of R33) is to develop an instrument that interacts with users by keyboard and performs the following biochemical steps automatically: (1) Polymerase chain assembly (PCA) of 1,000-2,000 base pairs (bp) double-stranded DNA fragments in each of two cuvettes in less than thirty minutes and with an accuracy of less than one error/1,000 bp. (2) Immunocapture of the 5BrU and 6mA hapten-labeled DNA fragments. (3) Treat the DNA fragments with T4 polymerase. In the absence of dNTP's, the 3'exonuclease activity of the T4 polymerase removes 20-30 nucleotides from the 3'ends of the DNA fragments. (4) Join the two DNA fragments, which contain 5'single stranded ends, following a ligase-independent cloning strategy. The deliverable at the end of the R21 phase would be a prototype that performs the first two steps. The de novo gene assembly technology proposed here allows genes to be synthesized on a time scale of a few hours once important proteins are identified. The synthetic genes can be cloned and overexpressed to provide a large quantity of pure protein. Purified proteins may be used to generate high-affinity reagents that can be employed in protein arrays. The development of protein arrays will provide a quick and efficient procedure to assay the expression and regulation of the protein under investigation and of other proteins involved in toxic-response. The combination of rapid automated gene synthesis with proteomics may accelerate the development of new disease markers and drug targets.